PROJECT SUMMARY The goal of this K01 Mentored Research Scientist Career Development Award is to provide the candidate with the conceptual knowledge and technical skills needed to pursue an independent research career as a computational cognitive neuroscientist focused on aging and late-life mood disorders. The candidate has a strong background in cognitive paradigm design and functional neuroimaging methodology, and an evolving interest in computational modeling of cognitive and affective dysfunctions in depressed older adults. The research study will use computational approaches to clarify disrupted mechanisms of reward and salience processes and may lead to findings that can serve as data-driven targets for future personalized treatments. The proposed training consists of formal courses, structured tutorials, and hands-on methodological instruction intended to strengthen the candidate's understanding and development of computational models for the study of Positive Valence Systems (PVS) abnormalities in aging and late-life depression. Her mentoring team consists of accomplished investigators who will provide guidance and training in neurobiological and computational approaches for the study of reward and salience abnormalities. The research study complements the candidate's training plan, as it focuses on PVS dysfunctions that may contribute to reward processing abnormalities in normal aging and late-life depression. It is based on the prediction that aging-related abnormalities in reward circuits interacting with neurobiological abnormalities of depression may alter reward expectancy and reward responsiveness, leading depressed older adults to assign greater affective salience to negative stimuli. Accordingly, the study proposes to investigate the impact of the effects of aging and depressive symptoms on prediction error encoding of affectively salient stimuli at three levels of analysis: circuits, behavior, and self-report. The participants will be older adults aged 60-85 years with major depression (N = 34, stratified into two levels of severity) or no history or presence of psychopathology (N = 34). The proposed study will use task-based fMRI and computational modeling to examine how the dynamic interaction of age and late-life depressive symptomatology influences neural network functions and the resulting reward and salience processing behaviors. This study and the studies to follow promise to identify personalized behavioral and neurobiological targets for much-needed interventions. This work is timely, as advances in cognitive remediation, brain stimulation, and targeted behavioral interventions are becoming increasingly capable of influencing selective neurobiological functions and associated behaviors.